In implementing tests by dynamically changing configurations of an apparatus (hereinafter, referred to as a “dynamic configuration change test”), the apparatus that is actually executing instructions (hereinafter, referred to as an “instruction execution apparatus”) logically disconnects a target apparatus, to which the instruction execution apparatus is accessing, and checks instruction operations for cases where the access to the target apparatus is disabled.
The instruction execution apparatus has used an operating system (OS) to repeatedly execute the dynamic configuration change tests, in conventional dynamic configuration change tests. More specifically, the OS executes a test program(s) and raises an exception(s) on the test program by entering an OS command(s) for logically disconnecting at least one of the target apparatuses from the instruction execution apparatus, in the conventional dynamic configuration change tests. Thereafter, the OS repeatedly executes the dynamic configuration change tests by entering an OS command(s) for the purpose of reconnection of the logically disconnected target apparatus and by restoring the target apparatus to a connected state.
Moreover, the following method has been discussed as well: Upon detection of an abnormal operation during executing the program(s), the OS restarts the program(s), and repeats restarting the program(s) until normal operation is resumed.
In the above-discussed method, the program restart operation is carried out for preventing unexpected failures by detecting a program run-away. In consequence, the above-discussed method is not intended for the dynamic configuration change tests.
In a conventional system that includes multiplexed target apparatuses and the instruction execution apparatus and is capable of achieving the dynamic configuration change tests, the OS in the instruction execution apparatus logically disconnects one of the target apparatuses while it continues processing on the rest of the target apparatus(es). Consequently, for the purpose of multiplexing target apparatuses, the OS needs to match data on the disconnected target apparatus to data on the rest of the target apparatuses when the logically disconnected target apparatus is reconnected to the instruction execution apparatus. Therefore, the OS replicates the data on the logically disconnected target apparatus to the rest of the target apparatus(es) when the logically disconnected target apparatus is reconnected to the instruction execution apparatus. In other words, the logically disconnected target apparatus is restored to as it was.
For holding the same data on all target apparatuses during this restoration processing or this replication processing, the dynamic configuration change test needs to be interrupted. In so doing, the amount of storage capacity of the multiplexed target apparatuses reaches several gigabytes, so that it takes a couple of minutes for the restoration processing or the replication processing.
For the above-discussed reason, the test is interrupted for a couple of minutes each time the test is carried out. This has caused a disadvantage that test efficiency is low in the conventional dynamic configuration change test.